Profile image for Levin Kuhlmann

Dr Levin Kuhlmann

Research Fellow


Levin Kuhlmann completed his PhD at Boston University in Cognitive and Neural Systems in 2007 and since then has worked as a researcher at The University of Melbourne and Swinburne University of Technology. He is currently Research Fellow in the Centre for Human Psychopharmacology at Swinburne University of Technology, Honourary Senior Research Fellow in the Department of Medicine - St. Vincent's Hospital at the University of Melbourne and Visiting Fellow in the Department of Biomedical Engineering at the University of Melbourne.

His research involves signal processing, control theory, machine learning and computational neuroscience applications to neural engineering, neuroimaging, anaesthesia, epilepsy and vision. He is interested in how the brain processes information at multiple-scales, from neuron to whole brain, in order to create our ability to consciously experience the world, and utilising such an understanding to engineer improved diagnostics, interventions and therapies for brain-related medicine.

(Personal home page:

Research interests

Biomedical science; Neuroscience

PhD candidate and honours supervision

Higher degrees by research

Accredited to supervise Masters & Doctoral students as Coordinating.

PhD topics and outlines

A depth of anaesthesia monitor for all anaesthetics: Automated depth of anaesthesia monitors do not work reliably for all anaesthetics which have different molecular modes of action. Help us to find an algorithm that can track the depth of anaesthesia for all, or most, anaesthetics and outperform existing approaches.

Advanced control theoretic observer techniques for neural mass model parameter estimation with real data: Our team have developed a set of deterministic observers for state and parameter estimation for a class of neural mass models. We have provided proofs of convergence of the state and parameter estimates to the true values under various noise conditions. This is something not generally guaranteed with stochastic filtering techniques. Help us to advance this control theoretic observer design.

Benchmark data-mining based epileptic seizure prediction/detection: Many seizure prediction/detection algorithms have been developed but most have only been evaluated on short data sets and therefore there is still uncertainty as to whether they work or not. Explore existing and new data-mining based approaches to seizure prediction/detection using the ultra-long-term Neurovista database

Computational modelling of the brain networks underlying anaesthesia: This project will investigate computational modelling of brain networks underlying anaesthesia derived from our electromagnetic imaging studies. This project will attempt to link the multiple-scales from the known molecular mechansisms of anaesthetics to the EEG and MEG signal, by creating a model that reproduces the observed EEG and MEG.

Create an observability framework for optimally determining the number and location of sensors used to track neural systems: The brain is complex and consists of many oscillatory populations of neurons interacting across a large network. We can only position a limited number of sensors/electrodes in the brain regions we wish to monitor. Therefore there are limits to the number of underlying oscillatory populations we can observe. Help us to design a framework to optimally position sensors.

Imaging the brain networks underlying anaesthesia:  Our research has shown evidence for a common parietal brain network being involved in anaesthetic-induced reductions in consciousness for different anaesthetics. However, to gain more certainty we need to study the effects of different anaesthetics on the same individuals. Help us to achieve this through the study of anesthetics using electromagnetic brain imaging and network analysis methods

Neural model-based depth of anaesthesia monitoring: explore different computational models of anaesthesia and the brain, and their ability to be applied to the tracking of anaesthetic brain states and inference of underlying physiological variables using stochastic filtering or other techniques. 

Neural model-based epileptic seizure prediction/detection: Apply stochastic filtering and determisitic observer techniques to state and parameter estimation of neural mass models for the purposes of advancing seizure detection and prediction methods using the one-of-a-kind human Neurovista dataset.

Scaling up stochastic filtering techniques for neural mass model parameter estimation: Help us to develop more efficient stochastic filtering approaches for neural mass models to enable high resolution inference of neurophysiological changes with high-density (intracranial) EEG arrays. This will have a broad impact across neurophysiology and neuroimaging in general.


Available to supervise honours students.

Honours topics and outlines

The basic electromagnetic signatures of anaesthetic-induced loss of consciousness.: Anaesthesia offers a controlled way to study consciousness. This project offers the opportunity to analyse electromagnetic (EEG/MEG) imaging data obtained during anaesthesia, in order to quantify the effects of anaesthesia on the brain and consciousness. 

Fields of Research

  • Neurosciences - 110900

Teaching areas

Biomedical science;Neuroscience;Model-Driven Engineering


Also published as: Kuhlmann, Levin; Kuhlmann, L.
This publication listing is provided by Swinburne Research Bank. If you are the owner of this profile, contact us to update.

Recent research grants awarded

  • 2017: Critical slowing in epilepsy *; NHMRC Project Grants
  • 2012: Characterising dynamical complexity in brain electromagnetic activity in generalized anxiety *; Barbara Dicker Brain Science grant

* Chief Investigator